parsonage in - a local nobleman's manor or a fortified dwelling ... Introduction KAARMA Most Estonian literature on history accepts a PARSONAGE IN viewpoint that after the year 1227 when SAAREMAA- had surrendered to the German-Danish crusa­ A LOCAL ders, Saaremaa managed to maintain greater in­ dependence than other counties (Maesalu et al NOBLEMAN'S 1995, 54£; Uluots 1937, 36, 43; see Tarvel 1983, MANOR OR 117££). This special status came to an end after A FORTIFIED the suppression of the Peasants' Revolt of St. Ge­ DWELLING OF orge's Night in 1343-1345 (see Mall 1998, 161£; Magi 2002a, 149££; Markus et al 2003, 11). A CONQUEROR?

The reasons for this so-called special status of AntonParn, Saaremaa have been characterised by new deve­ Erki Russow lopments in the social arrangements from the end of prehistoric times. An archaeologist Marika Magi (2002a, 134££, 153£) alleges that the burial traditions of the islanders reflect influences of Christianity since the 13th century. The same aut­ hor associates the early church buildings in Saa­ remaa with the manifestation of the local elite, who played a significant role in converting the islanders into Christianity (see Magi 2002a, 153£; 2004a, 33). She explains the independent adop­ tion of a Christian belief by the islanders by re­ ferring to the study of an archaeologist Jaak Mall (1998, 162-165, Abb. n which suggests that the Roman style elements of the early building sta­ ges of the oldest churches in Saaremaa - , Kaarma and Poide, may date from the pre-con­ quest period. Magi suggests that the location of the church was connected with the dwelling hou- 49 Anton Parn, Erki Russow

se or manor of a local aristocrat. 1bis opinion is supported by material from Scandinavia, which see correlation between a ma­ nor (big farm) and church being located close by (Magi 2002a, 154f; 2004a, 31f; Markus et al 2003, 21). A specific example of such development in Saaremaa may be considered the Kaarma church together with a stone building in the neighbourhood (Magi 2004a, 31f).

The present paper does not analyse the social relations of the islanders during the post-conquest period, nor any possible ex­ pressions of the vassal status of the local elite. We shall concen­ trate on a single source - the stone building that was located close by the Kaarma church and which according to similar ca­ ses in Scandinavia has been considered to be an early medieval (i.e. 13th century) manor belonging to a local nobleman. We dis­ cuss the construction time of the stone building that was loca­ ted at the Kaarma parsonage and the use of this building, for two reasons. First, scientists have only briefly dealt with deter­ mining the possible type of the Kaarma stone building, instead mainly interpretations have dominated so far. Second, the pre­ sent research results allow more precise dating of finds that were revealed during cleaning out the cellar 10 years ago. In this respect the building is a mirror that reflects the social rela­ tions of the islanders.

Overview of the area and the founding of the church

Kaarma stronghold, that was known as the administrative cen­ tre of pre-historic Saaremaa, is located about 100 meters away 1 from the church and the parsonage (Fig. 1) . The stronghold itself was situated by the River P6duste. The estuary of the Ri­ ver P6duste formed a narrow gulf, cutting deep into the inland of the island. It is supposed that an early port of the Kaarma 50 Kaarma parson<~g e in S<~<~rem<~<~ - <1 loc<1l nobl em<~n 's m<~nor or <1 fortified dwelling ... inhabitants was Fig. 1. Location of situated by the Kanrmn. gulC that pro­ Dmwing by Kersti Siitnn bably yet in pre­ historic times was moved to to the location of the later resi­ dence of the Oesel-Wiek Bis­ hop (Magi 2004b, 151). Nu- merous Late Iron Age stone cairns testify of that the Kaarma region was densely populated (Magi 1998, 153f; 2002b, 203). Ad­ ditionally, three hoards have been found in the area: coin ho­ ards from the lO th and beginning of the 12th century2 and the Pilla treasure of coins and silver ornaments from approximately 1200 (Molvogin 1994, 520-527; Leimus 2003, 150££). Judging by the density of coin hoards found, the island at the turn of the 12 th and 13th centuries has probably been a place where the Got­ land and/or German m erchants spent their winter (Leimus 2002a, 1599, Fig. 4).

Written sources first m ention Carmele (Kaarma) contract of divi­ ding Saaremaa, concluded between the Order of the Brothers of 3 the Sword , the Bishop of Riga and the town of Riga in the year 1234. The contract assigned the Kaarma lands to the town of Riga (LUB I: CXXXIX). In 1236, following the intervention from the legate William of Modena, the Bishop of Riga and the town of Riga returned the lands to the Oesel-Wiek Bishopric (LUB I: CXLV; see Mall1998, 158££). An interesting fact for us is also mentioning the Kaarma stronghold (really a "timber barrier" constructed in 51 Anton Parn, Erki Russow

Kaarma) in 1261 when it was conquered by the Order. It was precisely 'in this year when the troops of the Order came to Saaremaa in order to suppress the uprising that had started in 1260, after the Lithuanians had defeated the Germans in the Durbe battle in 1260 (Arbusow 1918, 46). A contract was conc­ luded between the islanders and the Master of the Order in Kaarma, like previously in 1241 and in 1255 (LUB 1: CCLXXXV).

Kaarma was the midst of the Oesel-Wiek bishopric's lands, and was most probably the ecclesiastical centre of Saaremaa (Mar­ kus et al 2003, 11). The importance of the church is stressed by its guardian saints Peter and Paulus, since mainly dome chur­ ches were devoted to these saints (Markus et al2003, 12). At the same time a correlation between churches and castles has been detected regarding the above mentioned saints in Pomerania (Heyden 1964, 59). Characterising the location of the Kaarma church (outside measurements 15.4 x 47.6 m) it is worth stres­ sing that it was built into a pre-historic centre. At the same time the location of the church in the landscape was not the best choice. The church was constructed on a natural sandy hill, the top of which was directly under the vestry and the South wall of the nave on the slope. The soft soil and the different depths of the basements started to create sinking problems already du­ ring the construction works. This is testified by numerous sup­ porting pillars (Markus et al2003, 30££). Three different theories exist as to the construction time of the church. An early opinion suggests that the church was founded in Kaarma after the re­ volt of the islanders in 1261, while the vaults of the nave collap­ sed already in the 13th century (Raam 1996, 58; Alttoa 2003, 33). Jaak Mall (1998, 163, 165) offers a theory that the church was established earlier, the oldest Romanesque style elements in the vestry may date from the pre-conquest time, i.e. from times before the year 1227. It is not credible in his opinion that the 52 Kaarma parsonage in Saaremaa - a local nobleman's manor or a fortified dwelling .. . church construction began in 1227-1236 or straight after the con­ quest and before the final establishment of the Oesel-Wiek bis­ hopric. Kersti Markus (2003, 87ff) suggests that the Kaarma church was built in mid- 13th century, in two different stages. The building must have been completed at least by the year 1260, and was then burnt down by the Order in 1261 (Markus et al 2003, 92ft). The fire caused the vaults to give excessive pres­ sure to the South and West wall, which partly sank and caused the nave vaults to collapse. The church was restored by the be­ ginning of the 1270s, first with a wooden ceiling.4 Markus also agrees with Marika Magi that the construction of stone chur­ ches in Saaremaa could only take place with direct involvement of the local elite. It is worth mentioning that the civil war bet­ ween the Order and the Bishop in 1298 is connected with the Kaarma church. The troops of the Order forced themselves into the Kaarma church and killed the people seeking refuge there (Saaremaa 1934, 500).

Research of the cellar of the parsonage's main building

The main building of the parsonage of the Kaarma church, which is the focus of the present article, is located about twenty me­ ters from the church. The parsonage belongs to one of the lar­ gest parsonage ensembles in Estonia. The main building of the parsonage was probably destroyed during the Livonian War in the last quarter of the 16th century. Archival material proves that the restoration of the parsonage started in the 20s of the 17th century (Markus et al2003, 22). The main building was again turned into ruins after the II World War, restoration work be­ gan only in 1991. There are various theories concerning the const­ ruction of the main building, too. Armin Tuulse (1942, 320f) is of the opinion that the original building at the site of the main 53 Anton Parn, Erki Russow

building of the parsonage was a manor, constructed as a tradi­ tional fortified dwelling house. The art historian dates the buil­ ding to the 16'h century. During the restoration works of the main building (total length 48 m) that were started in 1991 a two-room cellar beneath the main building was opened up to the full extent. Outside measurements of the cellar were 10.1 x 21.7 meters (Fig. 2). In a 1,7 m thick wall there is a 0.57 m wide stairway. The thickness of the cellar walls ranged from 1.27 to 1.4 meters. Originally there were three windows at the Nort­ hern side-wall and one window at both ends. Originally there Fig. 2. had been a beam ceiling, later reconstructions divided the Wes­ Ground tern room into two and vaulted ceiling were built. Later re­ plan of the Kaa rma constructions also added stairways to both cellars from outsi­ parsonage de. Research carried out in the cellars created two slightly dif­ cellar. Drawing by ferent results. T6nu Parmakson (1996, 59), who studied the const­ Tiinu ruction phases of the cellar, suggested that the cellar might ha­ Parmakson and Kersti ve belonged to a one- or two-storied fortified dwelling house Siitan from the first half of the 13'h century, that was most probably

54 Kaarma parsonage in Saaremaa - a local nobleman's manor or a fortified dwelling ... built for the church builders. T6nu Sepp (1995, 459££) who con­ ducted archaeological supervision in clearing process of the cel­ lar, noted two periods of use, which we shall discuss more pre­ cisely further on.

In conclusion, a significant amount of finds from the medieval and early modem times (SM 10108: 1-235) was collected during the constructional and archaeological research in the parsonage cellars in 1993. Analyses carried out during the past ten years have often been based on these finds, dating the older part of the parsonage to times prior to the construction of the church or at least to the same period (e.g. Markus et al 2003, 20-22 and reference literature there). Mainly a couple of the most signifi­ cant finds have been the basis for far-reaching results.

Re-evaluation the age of the finds

Closer examination proves that unfortunately it is not possible to use the finds for evaluating the exact date of the parsonage. The majority of finds were revealed from the early modem period dump and filling layer in the Eastern part of the cellar, hence it is not possible to connect them with absolute certainty with the acti­ vities in this specific building. However, we may still ascertain, having analysed the composition of the finds and their finding spot, that with great probability these objects come from the buil­ ding in question and/ or from the early building on the same spot. Most likely a greater part of the objects got into the cellar in the course of two Events (I and II). Since apart from the location plan we lack detailed information concerning the stratigraphic position of the finds, we can only rely on circumstantial evidence.

Event I is the destruction of the building sometimes during the 1 l6 h century. The destruction probably brought about also the 55 Anton Parn, Erki Russow

Fig. 3. collapse of the ground floor, and some of 11 Enrly 16 ' stove tiles the finds revealed during the research from tl1 e must have dropped into the cellar. Da­ pnrsonnge cellnr. ting this ev ent is fairly difficult. T6nu P/10 !o Erki Sepp, the archaeologist who directed the Russow work, dated the d estruction of the par­ sonage with a coin that was found there (Johan III Tallinn killing, 1570- 1580s) to the last quarter of the 16'h century, con­ necting it to the forays of the Livonian War (1558-1583) (Sepp 1995, 465). Later a more specific dating has been offered - the year 1576, when Russian cavalcades looted Saarem aa (Markus et al 2003, 22). Comparing this to other finds from the 16'h century, the connection of this event to the last decade of the 16tl' century se­ ems quite probable. Yet, the lack of de­ tailed information on the finds does not allow to totally rule out the probability that the building may have been destro­ yed at a somewhat earlier period. Name­ ly, the finds include some tiles and tile fragments with late Gothic influence from the beginning of the 16th century that are rather unique in Estonia (Fig. 3). Further­ more, earlier publications rely on the stam­ ping time of the Johan III Tallinn killing when determining the dating, at the sa­ me time overlooking the fact that such coins were comparatively widely used in Estonia even during the first quarter of 56 Kaarma parsonage in Saaremaa - a local nobleman's manor or a fortified dwelling ... the 17th century.5 Hence we cannot omit the possibility that the coin was left in the soil during the next event when reconstruc­ tion of the parsonage was started. So we may conclude that the exact time of the destruction of the parsonage is not determi­ ned with absolute certainty.

The other part of finds most probably came to the Eastern part of the parsonage cellar during the reconstruction of the buil­ ding (Event II). Since the parsonage cellars had narrow win­ dows and the only exit from the cellar was a narrow stairway (Fig. 2), we may presume that minimum labour was put into the restoration of the building and cleaning of the cellar. Probably some of the filling from the Western part of the cellar was re­ moved to the Eastern part, consequently finds from very diffe­ rent periods (13th- 17th centuries) ended up in a common filling and/or dump layer. It is not certain when exactly the cellar was filled. It is known that the restoration of the building started in the 1620s, major reconstructions also took place in 1732-1733 (Markus et al 2003, 22-23). We may say that dating the finds requires a lot of precaution. Yet it is worth stressing that despi­ te the above-mentioned difficulties, the Kaarma parsonage finds are noteworthy and deserve full attention. Kaarma parsonage finds are unique in the context of all Estonia, since no other stone building outside towns and castles has undergone such profound archaeological research. The finds also give us an over­ view about the material culture in the periphery.

Finds revisited

Kaarma finds have been thoroughly researched on several oc­ casions (e.g. Sepp 1995, 460-465; Pi.ii.ia 2000; Markus et al 2003, 20-22). Therefore we would like to concentrate on only a few

57 Anton Pam, Erki Russow

spectacular aspects that are directly connected to the presumable construction time in the 13th century.

Earlier researchers have based their datings of the building on the singular find that is precisely stratified- a bone comb (Fig. 4). The comb was discovered in the Southern corner of the Eastern cellar, from the soil beneath the basement walL The archaeologist who directed the excavation interpreted this soil as dating from before the building was constructed. Hence the find was in the soil before the stone building was established. Knowing the exact time of making and using the object allows us to date the time of the construction work

The comb that was found from beneath the parsonage walls belongs to the type of bone combs with concave end plates. Typologically these are medieval combs that were widely used in the whole Baltic Sea region (Luik 1998). Dating the Kaarma comb T. Sepp considered parallel finds from Viljandi and Tartu, where similar finds have previously been dated to the 11th- 13th centuries (Sepp 1996, 8). Later the proposed dating has deter­ mined that the building was constructed either prior to the es- tablishment of the church or to the sa­ me period, i.e. the first half or mid 13th century (Mar­ kus et al 2003, 20) . Fig. 4. Bone comb However, the aut­ revea led hors of the monog­ beneath the ba sement of raph have discar­ the ded the fact desc­ building . Photo Erki ribed in the archa­ Russow eological report, 58 Kaarma parsonage in Saaremaa - a local nobleman's manor or a fortified dwelling ... that such combs were made until the 15th century, and a recent research shows that combs with concave end plates were in use until the 16th century (Kallstrom 2002, 152).

The excavation report and the article about the research results (Sepp 1995; 1996) stipulate the finds from the parsonage cellar into two periods of use. The first period covers the time from the mid 13th century to mid 14th century, the second period co­ vers the 15th and 16th centuries (Sepp 1995, 465). This interpreta­ tion is largely based on imported pottery. Later analyses of the material dated the second period to ca 1550-1620 (Russow 2001, 115-117). There are only very few stray finds between the two periods. T. Sepp has interpreted the period from the second half of the 14th century until the beginning of the 15th century with very little finds as a period when the building was (al­ most) not used (Sepp 1995, 465). The authors of the present ar­ ticle consider that such interpreting of finds from the early mo­ dem period dump layer is not sufficiently grounded. Only finds from the first period are of interest to us when dating the buil­ ding. From these finds a metal object and a dozen pieces of pottery can be dated with accuracy.

From the numerous medieval metal objects (e.g. foot of a bron­ ze tripod pot, fragments of knives, locks etc.), it is possible to precisely date only the fragment of an armour plate that is bro­ ken into three pieces. In the research results the age of the ar­ mour plate is connected with the military conflicts during the second half of the 13th century (1260 or 1298) (Sepp 1995, 464). Later data suggests that the fragments date the armour rather to the end of the 13th century or even later.6 Therefore, these arguments cannot also be considered as convincing. In dating the building we should rather consider the ceramics find and first of all imported pottery. 59 Anton Pam, Erki Russow

Imported pottery of the parsonage from the first period consist of fragments of ten vessels. The oldest of them are fragments of a protostoneware jug from the Rhineland and nearstoneware jug made in Siegburg. Supported by parallel finds from the West Estonian towns (Russow 2001) and also considering the shape and side profile, the fragments of both jugs date from the last quarter of the 13'11 century (Fig. 5). To the same period, alt­ hough we cannot exclude later datings, belong fragments of an nearstoneware jug from Southern Lower Saxony and Langer­ wehe. From the typical imported ceramics of the first half of the14'11 century fragments of stoneware jug from Langerwehe and grey pottery from North Germany were also found from the parsonage cellar. The latest of the finds from the first pe­ riod is a brim fragment of stoneware covered with brown dap­ pled glazing from Southern Lower Saxony.

A bigger challenge is determining the time and origin of the weakly burnt local pottery. The excavation report divides them into two groups. A couple of fragments are thought to belong to the traditional ll'h - 13th century hand-made pottery (Sepp 1996, 3). Considering the exceptional shape (a pot with a spout - see Sepp 1995, Plate XXIV, 1), composition of the clay and d egree of bur­ Fig. 5. Near ning we cannot stonew11re agree with this d11 ting from the end of dating. The late 11 the 13 ' Iron Age and century Early Medieval from the parsonage (12'11 - 13'h centu­ cellar. ries) archaeologi­ Photo Erki Russow cal finds from Western Estonia lack similar ob- 60 Kaarma parsonage in Saaremaa - a local nobleman's manor or a fortified dwelling ...

jects. At the same time, such vessels of similar burning degree, finishing and clay composition were used in South Estonia du­ ring the 16th century. Also, pottery with similar brim profiles were used throughout the Middle Ages. The second group of local pottery is represented by a brim fragment of a tripod coo­ king pot with a higher burning degree. The design of the object is similar to pottery made in Tallinn in the second half of the 13th century and first half of the 14'h century (Mall & Russow 2004, 152-153, Abb. 5).

Summing up the analyses of finds from the 13th - 14'h centuries gathered during the archaeological research of the Kaarma par­ sonage we can conclude that the discussed finds do not include any finds that can be dated with absolute certainty to the mid 13 th century or an earlier period. For a certain number of finds (first and foremost the bone comb, knives) it is possible to de­ termine the time of their making and use to a period between the 11 th -13/ 14'11 centuries, but the composition of the finds still refers to typical households from the end of the 13th - begin­ ning of the 14th centuries. We have to stress, however, that the finds originate from a dump and filling layer and can therefore not be considered to be a basis for drawing larger generalisa­ tions. For the same reason the finds cannot be used as argu­ ments in dating the building. In conclusion we may say that dating the building beneath the parsonage requires further re­ search in future.

Cellar revisited

.Referring to the above mentioned, A. Tuulse (1942, 320f) loo­ ked upon the building that originally stood on the parsonage main building as a development of a typical 16th century forti­ fied dwelling house. T. Parmakson and T. Sepp consider that 61 Anton Parn, Erki Russow

the cellar belongs to a 13tl' century dwelling house without groun­ ding their theory. A more profound study of the original buil­ ding together with the cellar was carried out by K. Markus, who recognises great similarity in the masonry of the vestry (the oldest part of the cellar) and the cellar, comparing the choi­ ce of stones and their layout (Markus et al 2003, 20££; Fig. 17). The art historian doubts the earlier assumption that the buil­ ding was erected for the church builders, claiming that accor­ ding to the treaty with the islanders in the mid 13th century the fogt was allowed to visit the island once a year for law courts and there was no need for a separate building.7 Also, according to K. Markus the building could not have belonged to a priest as until the 14th and 15th centuries the priests did not own such massive buildings in Gotland. Referring to the Danish and Got­ land research results where medieval multi-stored buildings in the proximity of churches belonged to noblemen's families, K. Markus (2003, 21) found it probable that also the original buil­ ding in Kaarma might have belonged to a representative of the local elite (see Wienberg 1993, 152). The author agreed with the two separate periods of use of the building (second half of the 13'h century until the 14'h century and 15'h - 16'h centuries), but specified later reconstructions. Based on the details of the car­ ved stone found from the cellar, the first reconstruction might have taken place in the beginning of the 15'h century (Markus et al 2003, 21; Fig. 18). The demolished entry in the Northern wall of the cellar must date from the same period. Reconstructions were added also in the mid 15'h century and in the beginning of the 16th century. The late Gothic tiles found from the dump in the cellar date from the last renovation.

Taking into account the above mentioned arguments we are of the opinion that the typology of the original building still needs further discussion, mainly because no direct reference to any 62 Kaarma parsonage in Saaremaa - a local nobleman's manor or a fortified dwelling ... similar building type has been made in dating the construction time. We share an opinion that in dating the type of the buil­ ding the size of the cellar is of considerable importance (outside measurements 10.1 x 21.7 m), not omitting the spatial division of the cellar (Fig. 2). We should not overlook the fact that all research work concerning stone buildings from the 13th -14th centuries (with the exception of sacral buildings) in Estonia ha­ ve concentrated on town buildings. A brief look at the research results shows that since the mid 13th century cellars appear in joint buildings in the West Estonian towns like Lihula and Ha­ apsalu (Pam 2001, 599££) . In these cases there was a wooden dwelling house at the street with a stone half-cellar with a stair­ way behind it (e.g. measuring 5 m x 3.5 m; 5.2 m x 7 m). Helmi Oprus (1976, 152£, Fig. 11-13) has brought out a special type from the 14th century houses in Tallinn, where the fore-house by the street was wooden with a rectangular shaped stone buil­ ding behind it. The latter had a single room cellar in it. Neither of these examples fit the original house in Kaarma. From a neigh­ bouring area to Saaremaa - Gotland, we know of several stone storehouses built to the rural areas in the end of the 12th centu­ ry, and especially during the mid 13th century (Svahnstrom 1976, 13-16, Fig. 7, 8; Westholm 2001, 744££, Fig. 2, 3). An interesting example for us is the storehouse of the Bringes farm, dated 1240s in Norrland with its two-room division and a central stairway connecting the down- and upper floors. Unlike in Kaarma, the rooms are vaulted and the house does not have a cellar. Stone dwelling houses that are from the same period as the storehou­ ses in the countryside of Gotland and smaller than the original house in Kaarma (e.g. 8 m x 13m; 8 m x 14.5 m; Westholm 2001, 746-747, Fig. 4, 5). Neither can we find direct resemblance in Visby, where similar houses with the storehouses built to the countryside can be found, the so-called "storehouses without a clear dwelling function" (Westholm 748ff, Fig. 7, 9, 10). Their 63 Anton Pam, Erki Russow

measurements range between 7 m x 13 m; 8 m x 14.5 m; 5 m x 15.5 m. In conclusion we can say that a typical three-storied house type - a house "with ante-room, living space and ware­ house floor below barrel vaulting, and with loading hatces in the gable" (Westholm 2001, 759f) had developed in Gotland by the mid 13th century.

It seems that concentrating on searching for the so-called old original building and a type similar to it does not give us any results. At the same time a new house type- a fortified dwel­ ling house - is spreading in the end of the 15th century and the first half of the 16th century, known as the tower-house of local aristocracy (Albrecht 1995, 174 ff; Fig. 232, 233, 236). Their fore­ runner- rectangular small strongholds- are known from the 14th century Lower Saxony, Schleswig-Holstein and Denmark. A similar house type later developed from such small strong­ holds. Besides the local aristocracy also the clergy built such dwelling houses for themselves (Moller in print). These buil­ dings had a cellar, two main floors and an attic. Originally such houses were rather small, e.g. 6.8 m x 8.9 m; 7.1 m x 10.5 m. Normally each floor formed a single room.8 A few larger "for­ tified houses" stand out from this house type in Denmark and Sweden at the tum of the 16'h century (see Albrecht 1995, 185- 189), like for example:

Voergaard (Jutland; ca 1520); Tjele (Jutland; ca 1530s; 17m x 10m); 0strupgaard (Funen; 151 quarter of the 16'11 century); Bollerup (Skarie; 2n ct half of the 15u' century); Glimmingehus (Skane; 1499- 1505; 29m x 13m); Torpa (Vastergotland; end of the15'h century; ca 18m x 14m).

The most interesting from them are the "fortified houses" of Torpa and Glimmingehus, but especially Torpa (Albrecht 1995, 64 Kaanna parsonage in Saaremaa - a local nobleman's manor or a fortified dwelling .. .

Fig. 240, 24t 242). The Torpa "fortified house" especially has several similar elements with the Kaarma original house. The most significant of them are described below. The Torpa house also has a cellar divided into two parts, while one of the rooms (similarily to Kaarma the right-side room) has cylinder vaults. The stairway from the cellar to the ground floor is against the inner wall; dividing the cellar into two parts. Although in this case it is not a wall-stairway, we can see wall-stairways at the end walls of the second and fourth floor. The Torpa house too has massive stone walls (see Albrecht 1995, 187). The last exam­ ples testify that the Kaarma original house may have been const­ ructed later, between the end of the 15th century and beginning of the 16th century. During this period Saaremaa was under Da­ nish rule (1559 / 84-1613). Concerning the earlier finds from the dump layer in the cellar, then these may date from an even earlier building. We are of the opinion that a surf on the Nort­ hern outer wall of the parsonage refers to an earlier building at the Kaarma parsonage. In the bottom of the surf a burnt layer with heavily burnt granite stones was fixed (Sepp 1995, 460; Fig. 2).

Returning to the late medieval fortified dwelling house in Ka­ arma, we can say that this refers to the importance of the place. During this period Kaarma had a central role in Saaremaa, which is proved by the fact the church was designated as a dome church in 1450-1460s (Markus et al 2003, 101). At the time the bishopric was divided between two bishops, one had control over the islands Saaremaa and Hiiumaa and the other over main­ land Ui.anemaa. A stone building similar to the Kaarma house can be found in the neighbourhood- in Valipe, Hiiumaa. Valipe fortified manor is mentioned in the manuscripts in the year 1529. Research of the manor ruins determined the measurements to be 14.5 m x 28.7 m (Aluve 1996, 96) . 65 Anton Parn, Erki Russow

Conclusion

The present overview of the development of the Kaarma origi­ nal house shows that the conclusions drawn on the basis of this house about the development of the post-conquest society in Saaremaa during the first half of the 13th century, when the lea­ ding role was played by the local elite, require additional grounds. It seems that the examples of the Kaarma church and the stone building are not suitable for drawing such conclu­ sions. The Kaarma original house will certainly remain a valu­ able source, which may always be re-investigated.

Although the early dating of the original house plan is not cor­ rect, we find that the finds from the dump layer of the cellar from the end of the 13th century are of utmost importance. It is known that Poide stronghold was first mentioned in 1290 on the territory that belonged to the Order in Saaremaa. The earlier finds from Kaarma that we presented originate from approxi­ mately the same period. Although the finds are rather modest, still both reports together may refer to the so-called new condi­ tions on the island. Indirectly, similar processes on the main­ land can support this theory. In 1284 the Oesel-Wiek Bishop issued an order that mainland peasants have to pay tithe that in addition to the regular tenth the peasants had to provide a con­ siderable monetary fee (Leimus 2002b, 9). The 1294 town's right of Haapsalu, which was the administrative centre of the Oesel­ Wiek bishopric, limited the circle of inheritors for Estonians, restricting it to family members living in the town only (Pam 2004, 281). Whether these notices are correlated, needs to be discussed in the future, yet such ideas about changed condi­ tions during the second half of the 13th century came to mind when we reconsidered the research results of the cellar in the Kaarma parsonage. 66 Kaarma parsonage in Saaremaa - a local nobleman's manor or a fortified dwelling ...

References

The name Kaarma may have several interpretations: e.g. "kaarmad" - Engl. blackberries, "karma" - Engl. small bay, or "kaar" - Engl. arch, vault (Kallasmaa 1996, 63f). In the first case, the finding place is unknown; the second hoard comes from the Kaku village. In addition, lOth century Arabic coins have been found from the cairn (based on the oral information oflvar Leimus, numismatic of the Estonian History Museum). The Order of the Brothers of the Sword was destroyed in the battle of Saule in 1236, surviving members then joined the Livonian bra11ch of the German Order. The vaults were most probably built in the nave in the beginning of the 15th century (Markus ef a/ 2003, 98f). Oral information by Mauri Kiudsoo, numismatic of the Institute of History. Oral information from Jaak Mall, manager of the arms collection of the Estonian History Museum. The 1241 contract was renewed in 1255 and the court time for the fogt was set from Michaelmas Day (29. Sept) tmtil the Easter holidays, i.e. half a year (LUB I: CCLXXXV; Mall 1998, 160). One such example is known also from Muremoisse, Latvia (see Sparitis 1991, 283f; Fig. 9). Sources and manuscripts

LUB I = Liv-, Esth- und Curlandisches Urkundenbuch nebst Regesten. Bd. I. Hrsg. Fr. G. v. Bunge. Reval, 1853. Piliia, G. 2000. Kaarma kiriku pastoraadi. arheoloogiline leiumaterjal. Unpublis­ hed graduate thesis. Manuscript in the library of Chair of Archaeology of University of Tartu. Russow, E. 2001. Importkeraarnika Laane-Eestis ca 1225-1650. Unpublished M.A. thesis. Talfum; Tartu. Manuscript in the library of the University of Tartu. Sepp, T. 1996. Kaarma kiriku pastoraadihoone keldrite arheoloogiline jarelval­ ve. Kuressaare. Manuscript in the library of Museum of Saaremaa. Literature

Albrecht, U. 1995. Der Adelssitz im Mittelalter. Studi.en zum Verhaltnis von Archi.tektur und Lebensform in Nord- tmd Westeuropa. Miinchen. Berlin. Alttoa, K. 2003. Saaremaa kirikud. The churches on the island of Saaremaa. Tallinn. Aluve, K. 1996. Valipe (Vallipaa, Vallipea) kindlustatud m5isamaja. - Eesti Arhitektuur, 2, 96. Arbusow, L. sen. 1918. GrundriB der Geschi.chte Liv-, Est- und Kurlands. Riga. H eyden, H . 1964. Zum Schadegard-Problem. - Greifswald-Stralsunder Jahr­ buch, Bd. 4. Schwerin, 57-62. 67 Anton Parn, Erki Russow

Kallasmaa, M . 1996. Saaremaa kohanimed, 1. Tallinn. Kallstrom, M. 2002. Kammar och kamtillverkning i medeltidens Stockholm. - Upptaget. Arkeologi i Stockholm infor 2000-talet. Uppsala, 149-156. Leimus, I. 2002a. Millal ja kust tuli saksa kaupmees Liivimaale? - Akadeemia, 8, 1588- 1605. Leimus, I. 2002b. Saare-Laane piiskop - rikkaim mees Liivimaal? - Laanemaa Muuseumi Toimetised , VI. Haapsalu, 7- 12. Leimus, l. 2003. Kaarma Piila aardeleid 12. ja 13. sajandi vahetuselt. - Eesti Arheoloogia Ajakiri, 7, 2, 150-157. Luik, H . 1998. Muinas- ja keskaegsed luukammid Eestis. Muinasnja Tendus, 6. Tallinn. Markus et nl 2003 = Markus, K. , Kreem, T.-M ., Mand, A. 2003. Kaarma kirik. Eesti kirikud I. Tallinn. Molvogin, A. 1994. Die Flmde westeuropaischer Mi.inzen des 10. bis 12. Jahrhun­ d erts in Estland. - Numismatische Studien, Heft 10. Hamburg. Maesalu et al 1995 = Maesalu, A., Lukas, T. , Laur, M ., Tannberg, T. 1995. Eesti ajalugu , I. Tallinn. Magi, M. 1998. Districts and centres in Saaremaa. - Culture clash or compromi­ se? The Europeanisation of the Baltic Sea Area 1100-1400 AD. Papers of the Xl'h Visby Symposium held at Gotland C entre for Baltic Studies, Gotland University College, Visby October 41h- 9'h, 1996. Acta Visbyensia XI. Visby , 147- 157. Magi, M . 2002a. At the crossroads of space and time. Graves, changing society and ideology on Saaremaa (Osel), 9 1h- 13111 centuries A D . CCC papers: 6. Tallinn. Magi, M. 2002b. Piirkonnad ja keskused . Asustus muinasaja 15pu ja varakeska­ egsel Saaremaal arheoloogiliste, inimgeograafiliste ning ajalooliste allikate andmeil. - MT 11, 169- 232. Magi, M . 2004a. From Paganism to Christianity. Political changes and their reflection in the burial customs of l21h - 131h century Saaremaa. - D er Ostse­ eraum und Kontinentaleuropa 1100- 1600. EinfluBnahme-Rezeption-Wan­ del. CCC papers: VIIl. Hrsg. D. Kattinger, J. E. Olesen, H. Wernicke. Schwe­ rin, 27- 34. Magi, M. 2004b. " ... Ships are their main strength." Harbour sites, arable lands and chieftains on Saaremaa. - Estonian Journal of Archaeology, 8, 2, 128- 162. Mall, J. 1998. Verwaltungsgeschichte lmd Cluistianisierung der Insel bsel im 13.- 14. Jaluhlmdert. -Culture clash or compromise? The Europeanisation of the Baltic Sea Area 1100- 1400 AD. Papers of the XI'h Visby Symposium h eld at Gotland Centre for Baltic Studies, Gotland University College, Visby Octo­ 1 1 ber 4 h - 9 h, 1996. Acta Visbyensia XI. Vis by, 158-165. Mall, J. & Russow, E. 2004. Archaologie und Stalinismus - die Anfange d er Stadtarchaologie in Estland lmd die Ausgrablmgen auf dem Tallinner Rat­ hausplatz 1953. - Zeitscluift fiir Archaologie des Mittelalters, 2003/31. Borm , 145- 158. Moller, G. in print. Mittelalterliche Adelshofe in Stralslmd. - Castella Maris Bal­ tici, VII. In Print. 68 Kaarma parsonage in Saaremaa - a local nobleman's manor or a fortified dwelling .. .

Parmakson, T. 19 96. Kaarma p astoraat. - Eesti Arhitektuur, 2, 59. Parn, A . 2001. Dber die Hausbauentwicklung in Westestland im 13. bis 15. Jahrhundert. - M . Glaser (Hrsg.), Ltibecker Kollo­ quium zur Stadtarchaologie im Hanserau m III: Der Haus­ bau. Liibeck, 595-604. Parn, A . 2004. Kiilaehitiste jaljed Haapsalu varases linnaehi.tu­ ses. - About hillfort and town. Studies in H onour of Vilma Trununal. Research into ancient times, 14. Ed . A. Haak, E. Russow, A. Tvauri. Tallinn; Tartu, 269- 289. Raam, V. 1996. Kaarma Peetri ja Pauli kirik. - Eesti Arhitektuur, 2, 58- 59. Saaremaa 1934 = Saaremaa. Maat eaduslik, majanduslik ja aja­ looline kirjeldus. Tartu. Sepp, T. 1995. Investigations at Kaarma parsonage. - Procee­ dings of the Estonian Academy of Sciences. Humanities and Social Sciences 44, 4, 459- 466. Sparitis, 0. 199 1. Untersuchungen zur Formensprache d er Re­ naissance in der Architektur und bildenden Kunst Lettlands. - Beitrage zur Renaissance zwischen 1520 und 1570. M ateria­ lien zur Kunst- und Kulturgeschichte in Nord- und West­ deutschland, Bd. 2. Marburg, 273- 294. Svalmstrom, G. 1976. Hauser und HOfe d er handelstreibenden Bevi:ilkerung auf Gotland waluend d es Mitlelalters. - Hauser und Hofe im Ostseegebiet und im Norden vor 1500. Beitrage zm Geschichte und Soziologie des Wolmens. Acta Visbyensia V. Visby-symposiet for hi.sloriska vetenskaper 1974. Visby, 9-28. Tarvel, E. 1983. Zur Problematik der Bauernaufs tand e in Estla.nd irn Kontexte der Cluistianisierung tmd Kolonisation des Lan­ des. - Die Rolle der Ritterorden in der Christianisierung tmd Kolonisierung d es Ostseegebietes. Universitas Nicolai Coper­ nici. Ordines mi.litares. Colloquia Tonmensia Historic a I. Hrsg. Z. H . Nowak. Torun, 115- 124. Tuulse, A. 1942. Die Bmgen in Estland u.nd Lettland. Verhand­ lungen der Geleluten Estnischen Gesellschaft XXXIII. Dorpat. Uluots, J. 1937. Die Vertrage der Esten rnit den Fremden im XIII. Jahrhundert. Tartu. Dprus, H. 1976. Das W ohnhaus in Tallinn vor 1500. - Hauser tmd Hofe irn Ostseegebiet und im N orden vor 1500. Beitrage zur Geschichte und Soziologie des Wohnens. Acta Visbyensia V. Visby-symposiet for historiska vetenskaper 1974. Visby, 141- 164. Westholm, G. 2001. Visby and Gotland - medieval building de­ velopment. - M. Glaser (Hrsg.), Liibeck er Kolloquium zur Stadtarchii.ologie im Hanseraum III. Der H ausbau. Liibeck , 743-763. 69 Anton Parn, Erki Russow

Kaarma klebonija Siarne straipsnyje aptariarna reta kairniska vi­ Saaremaa saloje - dura:mZiq architektfua Estijoje - Kaarrna klebo­ vietos didiko dvaras nija Saarernoje. Autoriai diskutuoja ankstesn~ hi­ ar uzkariautojq potez~, kuri teigia, kad akrneniniai pastatai Ka­ i tvirtintas namas? arrnoje buvo pastatyti veliausiai XIII arnziuje vie­ tos didiko arba baznycios statytojlJ:. Sio straips­ Anton Parn, nio autoriai analizuoja sias prielaidas rerndarnie­ Erki Russow si naujais tyrirno rezultatais: radiniais ir pastatlJ: Santrauka tipais. IS naujo perziurirni duornenys leidzia in­ terpretuoti tiek laikotarpio, tiek pastato funkci­ jos atzvilgiais. Nauja analize leidzia rnanyti, kad akrnenines konstrukcijos pastatas buvo pastaty­ tas XV a. antrojoje puseje arba XVI a. pradzioje. XIII-XIV a. radiniai, rasti akrnenines konstrukci­ jos pastato rusyje, priklauso ankstesniarn pasta­ tui. AutorilJ: nuornone, Kaarrna radiniai, datuo­ jarni XIII a. antrqja puse, rasytiniai saltiniai, ap­ rasantys Poide pilies ikfuirnq rytineje Saarernos dalyje, rodo saloje vykusius socialinius pokycius, tvykusius po mkariavirnlJ:.

70